Americans Distrust Companies With Their Data — What it Means for Retailers

According to a recent Harris Poll, most Americans don’t trust anyone with their data. This includes companies they deal with regularly, like banks and retailers. And it’s no surprising. It seems like a common occurrence to see headlines reveal yet another retailer hit with a data breach. Americans are starting to feel that data breaches are an unfortunate but unavoidable fact of life. In fact, one in five Americans would prefer to accidentally break a bone than have their identity information stolen.

As consumers become more mobile and social, and have more choices about where they shop and buy, the result is an exploding "dataverse" that's unstoppable. This data is being used to fuel a retail revolution that can deliver better customer experiences — but it can also be used to ensure these experiences are safe from hacks and fraud.

Consumers Distrust POS Devices Most
There are more than 6.6 million point-of-sale (POS) devices in the U.S. When asked what devices were most at risk for fraud, the No. 1 answer was in-store POS devices (followed by smartphones and then desktop/laptop computers).

Machine learning now enables retailers to maintain hypergranular behavioral profiles of all the devices throughout their enterprise. This can be done at scale for hundreds of millions of devices. For example, a retailer with thousands of stores and tens of thousands of devices can monitor usage history from the last three years to four years to detect unusual behavior patterns caused by malware (think Target’s data breach).

Consumers Want Privacy Over Personalization
When asked if they would give companies more access to personal information (e.g., geolocation data) in return for better personalized marketing offers, more than four in five consumers said no. This is a conundrum for marketers. In order to deliver targeted offers that drive purchases, retailers require more personal information from consumers. For instance, using mobile geolocation, offers can be made as consumers walk into a store (not after they’ve checked out and departed).

Machine learning “offer engines” can break the impasse of this chicken-and-egg scenario by curating product recommendations without needing human intervention. For instance, computers can read digital user reviews and present book recommendations automatically. Much of consumers’ fears about data privacy are based on concerns about other people gaining access to their personal data. Since computers don’t gossip at the water cooler and divulge secrets, perhaps they can be trusted in scenarios where human operators would not be.

Consumers Trust Banks More
For every dollar that the retail sector spends on technology infrastructure, banks spend three. Perhaps this may contribute to consumer attitudes about the types of companies they trust. One in 100 consumers said they trusted companies such as big retailers with protecting their personal data, compared to one in four who trusted banks more.

Until recent years, large-scale machine learning was available only to the largest of organizations such as government agencies or multinationals. However, progress within the last decade — advances in data science and algorithms; quantum leap in computer processing power; free-falling cost of computer memory and storage; and emergence of the ubiquitous cloud infrastructure — has made machine learning affordable to businesses of any size. Small and midsize retailers can access machine learning risk models via e-commerce platforms such as Shopify and BigCommerce, while larger retailers can develop in-house machine learning systems running on commodity computers within weeks, not years.

If retailers want to gain and retain consumer trust, they need to step up and take a page out of the banks’ book: make customer data privacy a top priority, create strict internal controls to minimize fraud and data breach risk, and employ machine learning to stop threats like payment fraud and identity theft in their tracks.

Sandeep Grover is senior vice president of global e-commerce at Feedzai, a machine learning platform for fraud and risk.